Baca-Garcia et al (Reference Baca-Garcia, Perez-Rodriguez and Basurte-Villamor2007) highlight some of the important issues related to current nosological systems but other issues need consideration. They voice their concern that with such a high degree of diagnostic instability, the validity of epidemiological, clinical and pharmacological research is questionable. However, in most studies appropriate diagnostic schedules and interviews are used for assessment of patients and a high degree of diagnostic stability has been shown for patients assessed in this manner (Reference Tsuang, Woolson and CroweTsuang et al, 1981; Reference Schimmelmann, Conus and EdwardsSchimmelmann et al, 2005).

Baca-Garcia et al (Reference Baca-Garcia, Perez-Rodriguez and Basurte-Villamor2007) did not discuss factors such as the level of qualification and number of years of experience in psychiatry of the evaluators, whether the patients were evaluated by the same or different assessors at each visit, the place (i.e. in-patient, out-patient, emergency setting) of first contact, the mean duration of contact, etc., which can influence diagnostic stability. It is also not clear whether at each follow-up proper diagnostic evaluations of patients were performed before diagnosis was recorded.

Furthermore, diagnosis was recorded using ICD–9 codes, but clinicians were using the ICD–10 classification system and this might have lead to errors in conversions and reconversions. Although Baca-Garcia et al reported that clinicians entered one or two diagnoses at the time of evaluation, they have not presented any data regarding comorbidity. Furthermore, when we compare the ‘diagnosis received in at least 76% of evaluations’ the diagnostic stability in the emergency setting was more than in the out-patient setting for all disorders except eating disorders. This perhaps reflects the likelihood of the evaluators recording the previous diagnosis rather than doing a complete diagnostic evaluation in the emergency setting.

Baca-Garcia et al raise issues which are common in day-to-day practice and highlight the fact that the proper evaluation of the patient requires use of appropriate diagnostic schedules and obtaining information from all possible sources. It is inappropriate to conclude from the study that our diagnostic systems and all research based on this nosological system are flawed.